Tiger Data (creators of TimescaleDB)

ClickHouse buffers small writes and flushes on a 400ms timer. Right design for bulk analytics. Wrong design when a dozen microservices are logging in real time. Giuseppe Pollio built LogTide on TimescaleDB after running the benchmarks: - 14,200 inserts/sec vs. ClickHouse's 250 at 100-event batch - 0.47ms vs. MongoDB's 304ms on a filtered query at 100K logs - 6.2ms latency flat under 50 concurrent queries, 1K to 1M records Hypertable partitioning routes queries to the active chunk. Continuous aggregates make dashboard counts nearly free because the work is done at write time. https://lnkd.in/enHhD6Cs #TimescaleDB #Observability #LogManagement #OpenSource #ClickHouse

Meet the Tiger Data team heading to HANNOVER MESSE 2026! We’ll be in Hall 14, Booth L18 from April 20–24. If you’re working with Postgres, time-series data, or industrial analytics, we’d love to connect. Stop by to chat with the team, see what we’re building, and explore how we can help with your data challenges. Book a meeting in advance: https://lnkd.in/eJRd_Zug

Come say hi!

We’re heading to Barcelona! Stop by booth #2 at GrafanaCON to meet the Tiger Data team. We’re working with teams building observability on Postgres - helping them scale Grafana dashboards with time-series capabilities and columnar performance, all without adding new infrastructure. If you’re curious how others are doing it (or just want to talk shop), come say hello. https://lnkd.in/euBzcHC9

Hey open source developers! If you've built something with TimescaleDB and never got around to writing it up, this is for you! I'm launching the Community Member Spotlight Series: a new case study series for real projects built on open source TimescaleDB. Stories get published on the Tiger Data case studies page, the newsletter, and social. The bar isn't "enterprise customer with an approved quote." it's more like "did you build something worth talking about?" I'm looking for things like a before/after query performance story, a sensor pipeline for a homelab or production IoT setup, a decision to stick with Postgres instead of splitting into two systems, or maybe an unexpected use case we'd raise an eyebrow at? Here's how it works: fill out a short form with 8 questions (10-15 min). I may follow up with a question or two. I write the draft, you review it before anything goes live, and everyone selected gets free Tiger Data swag! No Tiger Cloud requirement, no paid relationship needed. Open source only 😎 If you've been sitting on a good data story, this is your opening. Fill out the form and I'll do the rest. https://lnkd.in/eXiAJtZw #TimescaleDB #PostgreSQL #OpenSource #Developers #CommunitySpotlight

pg_textsearch 1.0 is now generally available. Also in this edition: ▪ Why peak throughput benchmarks miss what production actually needs ▪ What continuous ingestion does to Postgres maintenance windows ▪ Why hardware upgrades won't fix your IIoT throughput — the bottleneck is I/O, not compute ▪ VesselAPI moving 700K AIS position reports per hour from MongoDB to TimescaleDB ▪ Cactos cutting database costs by 55% after migrating from RDS to Tiger Cloud We're also at HANNOVER MESSE and AWS Summit London this month. Come find us. Read the full edition: https://lnkd.in/epMzy489

Battery energy storage moves fast. Database bills shouldn't. Cactos builds BESS platforms that protect agricultural and EV logistics operators from demand spikes that can 4x electricity costs for an entire quarter. On Amazon RDS, keeping that data fast and queryable was getting expensive. After migrating to Tiger Cloud, they unified relational and time-series data in a single PostgreSQL database. Results came in under a month: - 92% storage compression - Monthly costs cut by 55% "It makes things much simpler to manage everything from a single database,” said Juuso Mäyränen, Cactos Co-Founder & Software Engineer. One database. Full SQL. No second system. Read the full story: https://lnkd.in/ezY6cHdg #TimescaleDB #PostgreSQL #EnergyStorage #BatteryStorage #CloudDatabase

Tiger Data is heading to Hannover Messe 2026! We'll be at Hall 14, Booth L18 from April 20-24 in Hannover. If you're building on Postgres and running industrial telemetry, IoT sensor data, or operational analytics at scale, come find us. We'll show you how customers like Axpo, Waterbridge, and Flogistix handle billions of data points a day without splitting their architecture into a second system. Book a meeting: https://lnkd.in/eJRd_Zug

pg_textsearch v1.0 is now generally available, both on Tiger Cloud and as open source. If you’ve used Postgres’s built-in ts_rank for full-text search at scale, you’ve likely run into its limitations: no inverse document frequency, no term frequency saturation, and no efficient top-k path. Most teams work around this by bolting on Elasticsearch as a sidecar. That works, but now you're syncing data between two systems, operating two clusters, and debugging consistency issues when they diverge. pg_textsearch takes a different approach. It implements a full BM25 search engine directly inside Postgres. Tokenization, indexing, compression, and query execution are all built in C on top of Postgres’s storage layer. Indexes live in standard Postgres pages, managed by the buffer_cache. It participates in WAL, works with pg_dump and streaming replication, and requires no external storage. The engineering details matter here, especially for performance and scalability. We implemented Block-Max WAND with WAND pivot selection for fast top-k queries. Posting lists use SIMD-accelerated bitpack decoding. Parallel index builds handle 138M documents in under 18 minutes. On the MS-MARCO benchmark at that scale, pg_textsearch is 2.4x to 6.5x faster than ParadeDB/Tantivy for 2-4 term queries, and sustains 8.7x higher concurrent throughput. There are admittedly some tradeoffs. ParadeDB can build indexes roughly 2x faster. pg_textsearch does not yet support phrase queries, and the systems converge at 7+ lexeme queries. TJ’s blog post covers the full architecture, query optimization strategy, and benchmark methodology: https://lnkd.in/eE3ScyGZ Tiger Data (creators of TimescaleDB) #PostgreSQL #FullTextSearch #BM25 #OpenSource #TimescaleDB